Polar Biology

, Volume 33, Issue 1, pp 71–83

Culturable bacteria associated with Antarctic lichens: affiliation and psychrotolerance

Authors

  • Laura Selbmann
    • DECOSUniversità degli Studi della Tuscia
    • DECOSUniversità degli Studi della Tuscia
  • Serena Ruisi
    • DECOSUniversità degli Studi della Tuscia
  • Martin Grube
    • Institute of Plant SciencesKarl-Franzens-University Graz
  • Massimiliano Cardinale
    • Institute for Environmental BiotechnologyGraz University of Technology
  • Silvano Onofri
    • DECOSUniversità degli Studi della Tuscia
Original Paper

DOI: 10.1007/s00300-009-0686-2

Cite this article as:
Selbmann, L., Zucconi, L., Ruisi, S. et al. Polar Biol (2010) 33: 71. doi:10.1007/s00300-009-0686-2

Abstract

Antarctic habitats harbour yet unexplored niches for microbial communities. Among these, lichen symbioses are very long-living and stable microenvironments for bacterial colonization. In this work, we present a first assessment of the culturable fraction of bacteria associated with Antarctic lichens. A phylogenetic analysis based on 16S rRNA gene sequence of 30 bacterial strains isolated from five epilithic lichens belonging to four species (Lecanora fuscobrunnea, Umbilicaria decussata, Usnea antarctica, Xanthoria elegans) shows that these represent the main bacterial lineages Actinobacteria, Firmicutes, Proteobacteria and Deinococcus-Thermus. Within the Actinomycetales, two strains group in the genera Arthrobacter and Knoellia, respectively. Most of the other Actinobacteria form well-supported groups, but could be assigned with certainty only at the family level, and one is in isolated position in the Mycobacteriaceae. The strains in Firmicutes and Proteobacteria belong to the genera Paenibacillus,Bacillus and Pseudomonas, which were already reported from lichen thalli. Some genera such as Burkholderia and Azotobacter, reported in the literature as also associated with lichens, have not been detected in this study. One strain represents the first record of Deinococcus in epilithic lichens; it is related to the species Deinococcus alpinitundrae from Alpine environments and may represent a new species. Further separated and well-supported clades indicate the presence of possibly new entities. Some of the examined strains are related to known psychrophilic bacteria isolated from ice and other extreme environments, others with bacteria distributed worldwide even in temperate climates. Most of the strains tested were able to grow at low temperatures, but tolerated a wider range of temperature. Ecological and evolutionary implications of these lichen-associated bacteria are discussed.

Keywords

AntarcticaBacteriaExtreme conditionsLichensMicrobial associationsSSU rDNA

Introduction

The symbiotic association of fungi with photosynthetic organisms had a profound impact on the eukaryotic evolution. Mycorrhizal symbioses with plants likely supported the early colonization of land, the evolution of higher plants and vegetation as we know it today from temperate to tropical regions (Pirozynsky and Malloch 1975; Heckman et al. 2001; Schüßler et al. 2001). On the other hand, radiation of the fungi apparently correlates with the evolution of the lichen symbiosis of fungi and algae and/or cyanobacteria (Gargas et al. 1995; Lutzoni et al. 2001; Liu and Hall 2004; Reeb et al. 2004), some 600 million years ago (Yuan et al., 2005). The worldwide distribution of lichens and their estimated cover of more than 8% of land surface is a clear testimony to the ecological success of this lifestyle (Ahmadjian 1995). Other than higher plants, lichens are particularly robust in cold habitats, i.e. at high altitudes and latitudes. Under such ecological conditions, lichens are often conspicuous as the pioneering species. It is generally assumed that the extreme stress tolerance of lichens is conferred by the physiological integration of the symbionts and the ability to survive adverse conditions when desiccated, based on their ability to revive from the dry, anhydrobiotic state (Kranner et al. 2008). Even in the McMurdo Dry Valleys, Victoria Land, Antarctica, the most inhospitable terrestrial ice-free environment on earth, lichen crusts colonize rock surfaces in protected niches (Nienow and Friedmann 1993). There, at more exposed sites, lichens can also escape the prohibitive external conditions by hiding inside the air-spaces of sedimentary rocks, where the temperatures are buffered by the thermal inertia of the rock substratum. Hiding inside rocks, the thalli give up a more complex heteromerous organization that normally develops under less hostile conditions (De los Rios et al. 2005). Rather, they grow around rock grains as poorly differentiated consortia of fungal hyphae and unicellular algae, usually co-occurring with other eukaryotic and prokaryotic microorganisms. The arrangement of microorganisms inside the rocks, as cryptoendolithic communities, is macroscopically displayed as a coloured stratification (Friedmann 1982). Molecular studies of environmental DNA extracts from these communities suggested identity of an endolithic fungal phylotype with adjacent epilithically growing Buellia spp. (>97% similarity of SSU rDNA sequences; de la Torre et al. 2003). The same authors highlighted the microbial biodiversity in these communities comprising lichen mycobionts, free-living fungi, algae, as well as Actinobacteria, Alphaproteobacteria, Gammaproteobacteria and some other unidentified bacterial phylotypes.

As self-supporting and long-living organisms, lichens constitute a rather stable niche for other microbes. The presence of bacteria associated with lichens is known since a long time. Earlier, determination was based on phenotypical and physiological characters alone (Henkel and Yuzhakova 1936; Henkel and Plotnikova 1973), but recent molecular data from lichen-associated bacteria in temperate climates showed high bacterial biodiversity and abundance associated with lichens (Cardinale et al. 2006, 2008; Grube et al. 2009). The potential roles of lichen-associated bacteria could be manifold, but experimental tests with natural lichens are still pending.

Very little is still known about the biogeographic aspects of lichen-associated bacterial communities. Actinomycetes have specifically been isolated from lichens by Gonzáles et al. (2005), both from warm (Hawaii and Reunion Islands) and in cold climates (Alaska). This study indicated that a generally lower diversity is present in the sub-polar habitats. However, neither the lichens were identified nor were the involved bacterial strains characterized in detail. It thus remains unclear, whether the strains isolated from high latitudes comprised specifically adapted bacteria to cold climates. We hypothesize that lichens can enrich bacteria from their environments in different climatic zones, if these are able to colonize the hyphal structures of lichens or the lichen-influenced substrate interface. Bacteria, which can grow on organic substrates in cold environments, might include specialists that have evolved specific adaptations, some of which could also be of technical interest, such as by their production of enzymes with low temperature activity.

In this study, we investigated the diversity of culturable bacteria associated with Antarctic lichens. We focused on epilithic lichens from the most inhospitable continent, including its harshest part, the McMurdo Dry Valleys, to evaluate the possible role of lichens as a microbial niche as well as a source of new species in extreme conditions. Sampling in Northern and Southern Victoria Land, Antarctica was carried out in the framework of the Italian expedition that took place in 2003/2004, which was joined by one of us (Laura Zucconi).

Materials and methods

Sample collection

We analysed 16 epilithic lichens belonging to 13 species collected in different locations of Northern and Southern Victoria Land, Antarctica (Table 1). Samples were collected using a sterile chisel and preserved in sterile bags at −20°C.
Table 1

Lichen samples studied

Sample number

Lichen species

Location

Sampling date

1

Acarospora sp.

Ford Peak, Northern Victoria Land, 75°41′26.3″S 160°26′25.3″E

28/01/2004

2

Acarospora flavocordia Castello & Nimis

Kay Island, Northern Victoria Land, 75°04′13.7″S 165°19′02.0″E

30/01/2004

3

Buellia frigida Darb.

Inexpressible Island, Northern Victoria Land, 75°52′23.2″S 163°42′16.5″E

17/01/2004

4

Lecanora fuscobrunnea Dodge & Baker

Edmonson Point, Northern Victoria Land, 74°19′43.7″S 165°08′00.7″E

29/01/2004

5

Lecanora fuscobrunnea Dodge & Baker

Convoy Range, Southern Victoria Land, 76°54′33.0″S 160°50′00.0″E

25/01/2004

6

Lecanora sp.

Inexpressible Island, Northern Victoria Land, 75°52′23.2″S 163°42′16.5″E

17/01/2004

7

Lecideacancriformis Dodge & Baker

Widowmaker Pass, Northern Victoria Land, 74°55′23.5″S 162°24′17.0″E

12/02/2004

8

Rhyzocarpon sp.

Vegetation Island, Northern Victoria Land, 74°47′05.2″S 163°38′40.3″E

16/01/2004

9

Umbilicaria aprina Nyl.

Kay Island, Northern Victoria Land, 75°04′13.7″S, 165°19′02.0″E

30/01/2004

10

Umbilicaria decussata (Vill.) Zahlbr.

Kay Island, Northern Victoria Land, 75°04′13.7″S, 165°19′02.0″E

02/02/2004

11

Umbilicaria decussata (Vill.) Zahlbr.

Vegetation Island, Northern Victoria Land, 74°47′05.2″S, 163°38′40.3″E

16/01/2004

12

Usnea antarctica Du Rietz

Kay Island, Northern Victoria Land, 75°04′13.7″S, 165°19′02.0″E

30/01/2004

13

Usnea antarctica Du Rietz

Vegetation Island, Northern Victoria Land, 74°47′05.2″S, 163°38′40.3″E

16/01/2004

14

Xanthoria elegans (Link) th. Fr.

Kay Island, Northern Victoria Land, 75°04′13.7″S, 165°19′02.0″E

30/01/2004

15

Lecidea sp.

Starr Nunatak, Northern Victoria Land, 75°53′55.7″S, 162°35′31.3″E

15/02/2004

16

Lecidea sp.

Starr Nunatak, Northern Victoria Land, 75°53′55.7″S, 162°35′31.3″E

15/02/2004

Bacterial isolation

For the isolation of the external bacteria, c. 200 μl volume of lichen thalli were first vortexed for 1 min in 0.8% NaCl and 100 μl of the resulting solution was plated on tryptone-yeast extract (TY) medium, amended with Nipagin 0.03% w/v (methyl-para-hydroxy-benzoate, Sigma-Aldrich, Steinheim) to prevent fungal and algal growth. The isolation of the internal bacterial portion was performed as follows: lichen thalli were first washed 1 min in 0.8% NaCl and then sterilized in H2O2 (8%) for 5 min. Samples were then washed for 4 min in NaCl (0.8%) to remove H2O2 traces and crumbled in 500 μl of sterile deionized water using a sterile chisel. Finally, 100 μl of the resulting suspension was plated as reported above.

Plates were incubated at 10 and 25°C and inspected every 15 days until no new colonies appeared. For each lichen sample, all the colonies with distinctive phenotypes were purified by seeding single colony on new plate filled with the same medium.

Temperature relations

Temperature preferences were tested by seeding the isolated bacterial strains on Petri dishes containing TY medium and incubating at 0 ± 1, 5 ± 1, 10 ± 1, 15 ± 1, 20 ± 1, 25 ± 1, 30 ± 1, 35 ± 1 and 40 ± 1°C. Test was performed in triplicate. Plates were inspected weekly for 2 months.

Molecular analysis and phylotypes identification

Bacterial genomic DNAs were extracted using thermal shock procedure performed as follows: an overnight single colony was transferred through a micropipette into a 0.5-μl tube and suspended in 14 μl of sterile deionized water. The suspension was heated 5 min at 100°C and immediately cooled in ice. The suspension was centrifuged at 4,000g for 3 min and the supernatant utilized for the following PCR reaction. The amplification of the 16S rRNA gene sequence was performed using BioMix (BioLine GmbH, Luckenwalde, Germany) and the primers fD1 (5′-CCG AAT TCG TCG ACA ACA GAG TTT GAT CCT GGC TCA G-3′) and rD1 (5′-CCC GGG ATC CAA GCT TAA GGA GGT GAT CCA GCC-3′) (Weisburg et al. 1991). The amplification was carried out using MiniCycler™ (MJ Research, Waltham, MA, USA) equipped with a heated lid. The first denaturation step at 95°C for 5 min was followed by: denaturation at 95°C for 45 s, annealing at 55°C for 1 min and extension at 72°C for 90 s. The last three steps were repeated 35 times, with a last extension at 72°C for 5 min. The PCR products were visualized on agarose gel electrophoresis and quantified by comparison with the Ladder GeneRuler™ 1 kb DNA (Fermentas, Vilnius, Lithuania). The products were purified using Nucleospin Extract kit (Macherey-Nagel, Düren, Germany). Sequencing reactions were performed according to the dideoxynucleotide method (Sanger et al. 1977) using the TF Big Dye Terminator 1.1 RR kit (Applied Biosystems, Forster City, CA, USA). Fragments were analysed using an ABI 310 Genetic Analyser (Applied Biosystems). Sequence assembly was done using the software Chromas (version 1.45 1996–1998, Conor McCarthy School of Health Science, Griffith University, Southport, QLD, Australia). Sequences with high similarity available in NCBI GenBank were identified using BLASTn search (Altschul et al. 1997, http://www.ncbu.nml.nih.gov/blast/) and used for the alignments.

Alignment and tree reconstruction

Alignment was first carried out using ClustalX (Thompson et al. 1997), then exported to MEGA4 (Tamura et al. 2007) and improved manually. Due to gaps necessary for alignment, the 16S domain spanned 1,519 positions. The alignments were based on the positions 1–1432, the final parts were cut off to compare fragments with the same length. Alignments were then exported and the best-fit substitution model was determined using MrAic.pl 1.4.3 (Nylander 2004 program distributed by the author) estimated using Phyml (Guindon and Gascuel 2003) through hierarchical likelihood ratio tests. MrAic calculates the Akaike Information Criterion, corrected Akaike Information Criterion and Bayesian Information Criterion; Akaike weights for nucleotide substitution model and model uncertainty. All 56 models implemented in Modeltest were evaluated. Phylogenetic trees were reconstructed by Maximum Likelihood, using Treefinder (Jobb et al. 2004) and the resulting tree was displayed using Treeview 1.6.6 (Page 1996). The robustness of the phylogenetic inference was estimated using the bootstrap method (Felsenstein 1985) with 1,000 pseudoreplicates generated and analysed with Treefinder.

Results

Bacterial isolation

Thirty morphologically distinct bacterial strains (Table 2) were isolated from 5 lichen thalli (sample numbers 5, 10, 11, 12, 14) belonging to four species: Lecanora fuscobrunnea Dodge and Baker, Umbilicaria decussata (Vill.) Zahlbr., Usnea antarctica Du Rietz, Xanthoria elegans (Link) th. Fr., while no isolate was obtained from the remaining 11 lichen thalli. As much as 20 of the isolates were from the external surface and 10 from the inner part of the thallus. The number of different phenotypes isolated from each thallus ranged from 1 to 12. For most of the different bacterial phenotypes, very few colonies were obtained in the isolation plate. The number of colonies ranged from 1 to 5 for 22 phenotypes, between 10 and 50 for other 5 phenotypes, more than 50 for 1 isolate from U. decussata (10Ude1), and more than 100 for isolates 10Ude2 and 10Ude6 from the same lichen thallus.
Table 2

Culturable bacteria isolated from lichen samples

Bacterial isolate

Lichen species

Localization

Number of colonies

Most related sequences (GenBank accession number)

16S rRNA gene sequence similarity in %

Taxonomical position of isolate

Notes/references of most related sequences

5Lfe1

Lecanora fuscobrunnea

Ext

1

Saxeibacter lacteus (AM778124)

98

Actinobacteria; Actinomycetales

Isolated from a rock in Korea (Lee et al. 2008)

5Lfe2

Ext

1

5Lfe3

Ext

1

5Lfe4

Ext

2

12Uae5

Usnea antarctica

Ext

4

12Uae4

Usnea antarctica

Ext

5

Pseudomonas lutea (AY364537)

98

Gammaproteobacteria; Pseudomonadaceae; Pseudomonas

Soil bacterium from Spain (Peix et al. 2004)

12Uai1

Int

10

12Uae3

Usnea antarctica

Ext

1

Arthrobacter sp. (AM419018)

98

Actinobacteria; Actinomycetales; Micrococcaceae; Arthrobacter

Isolated from Antarctica

12Uae1

Usnea antarctica

Ext

1

Uncultured bacterium (AM696982)

98

Actinobacteria

Isolated from indoor environment

12Uae2

Ext

2

10Ude8

Umbilicaria decussata

Ext

1

10Ude1

Umbilicaria decussata

Ext

>50

Endophytic bacterium (DQ339615)

Uncultured bacterium (AM696980)

98

Actinobacteria

Isolated from alpine subnival plants (DQ339615)

Isolated from indoor environment (AM696980)

10Ude2

Ext

>100

10Ude4

Ext

25

10Ude5

Ext

1

10Ude6

Ext

>100

10Udi1

Int

5

10Udi2

Int

20

10Udi3

Int

2

10Ude7

Ext

1

10Ude3

Umbilicaria decussata

Ext

1

Deinococcus alpinitundrae ME 04-04-52 (EF635408)

97

Deinococcus-Thermus; Deinococcales; Deinococcaceae; Deinococcus

Psychrophilic, ionizing radiation-sensitive species isolated from an alpine environment (Callegan et al. 2008)

11Udi1

Umbilicaria decussata

Int

5

Paenibacillus sp. KM8 (AJ011321)

97

Firmicutes; Bacillales; Paenibacillaceae; Paenibacillus

Probably yet undescribed soil species of Paenibacillus from southern Finland (Elo et al. 2001)

10Udi4

Umbilicaria decussata

Int

1

Knoellia sp. (DQ812538)

99

Actinobacteria; Actinomycetales; Intrasporangiaceae; Knoellia

Marine bacterium from Antarctica

14Xee2

Xanthoria elegans

Ext

5

Pseudomonaslini (AY035996)

99

Gammaproteobacteria; Pseudomonadaceae; Pseudomonas

Isolated from bulk and rhizospheric soils (Delorme et al. 2002)

14Xee3

Ext

15

14Xee1

Xanthoria elegans

Ext

30

Bacillus sp. (AY553105)

99

Firmicutes; Bacillales; Bacillaceae; Bacillus

Halotolerant Bacillus species from the Great Salt Plains of Oklahoma (Caton et al. 2004)

14Xei4

Xanthoria elegans

Int

5

Mycobacterium sp. (DQ372729)

99

Actinobacteria; Actinomycetales; Mycobacteriaceae; Mycobacterium

A new soil Mycobacterium species isolated from an illegal dumping site in Japan (Wang et al. 2006)

14Xei2

Xanthoria elegans

Int

1

Paenibacillus sp. (EU558282)

99

Firmicutes; Bacillales; Paenibacillaceae; Paenibacillus

Psychrotolerant bacterium from Alaskan tundra

14Xei1

Xanthoria elegans

Int

1

Bacillussimplex (DQ275175)

99

Firmicutes; Bacillales; Bacillaceae; Bacillus

Isolated from granite rock, Santa Catalina Mountains, Arizon, USA

14Xei3

Xanthoria elegans

Int

5

Pseudomonas sp. (EF157292) Glacial ice bacterium (AF479376)

99

Gammaproteobacteria; Pseudomonadaceae; Pseudomonas

Isolated from a trinitrotoluene contaminated soil (EF157292) Isolated from glacial environment (AF479376) (Christner 2002)

Bacterial isolates included in the tree are reported in bold

Temperature relations

Data regarding temperature preferences of the bacterial strains tested are reported in Table 3. The strains 11Udi1 isolated from sample 11 of U. decussata, and 14Xei3 and 14Xei4 from X. elegans were not tested. Most of the strains tested were able to grow only at relatively low temperatures. For instance, 12 strains were able to grow only when incubated at temperatures from 15 to 25°C, with an optimum around 20–25°C. Ten strains were able to grow at 0°C. Most of the strains tested were not able to grow at temperature above 25°C with the exception of 10Udi4, 14Xei1, 14Xei2, 14Xee1, 14Xee2 and 14Xee3. The ability of the strains from X. elegans to grow very well even at 35°C is remarkable.
Table 3

Temperature preferences

Bacterial strain

Lichen species

Temperature °C

0

5

10

15

20

25

30

35

40

10Udi4

Umbilicaria decussata

+

+

+

++

+++

+

10Ude3

+

+

+++

+++

++

10Ude1

+

+

++

10Ude2

+

+

++

10Ude4

+

+

++

10Ude5

+

+

++

10Ude6

+

+

++

10Udi1

+

+

++

10Udi2

+

+

++

10Udi3

+

+

++

10Ude7

+

++

++

10Ude8

+

++

++

12Uae3

Usnea antarctica

+

++

+++

+++

+++

+++

12Uae4

+

++

++

++

++

++

12Uai1

+

++

++

++

++

++

12Uae5

+

+

++

+++

++

12Uae1

+

++

++

12Uae2

+

++

++

5Lfe1

Lecanora fuscobrunnea

+

+

+

++

++

++

5Lfe2

+

+

+

++

++

++

5Lfe3

+

+

+

++

++

++

5Lfe4

+

+

+

++

++

++

14Xee2

Xanthoria elegans

++

+++

+++

+++

+++

+++

+++

++

14Xei2

++

+++

+++

+++

+++

+++

+++

++

14Xee3

++

+++

+++

+++

+++

+++

+++

++

14Xei1

+++

+++

+++

+++

+++

+++

++

14Xee1

+++

+++

+++

++

Phylotypes identification and phylogeny

All 16S rRNA gene sequences matched with entries in GenBank with similarities ranging from 97 to 99% (Table 2). All the 16S rRNA gene sequences of the isolates from L. fuscobrunnea were identical, except for 5Lfe1, which was different in three positions only; all these four isolates, as well as the strain 12Uae5 from U. antarctica, were 98% similar with the novel actinomycete genus and species Saxeibacter lacteus isolated from rock (Lee et al. 2008). Five additional strains were isolated from the thallus of U. antarctica: 12Uae1 and 12Uae2 showed 98% similarity with an uncultured, unknown bacterium from indoor (AM696982); strain 12Uae3 showed 98% similarity with an Arthrobacter species (AM419018) isolated from Antarctica; 12Uae4 and 12Uai1 were both 98% similar in the 16S rRNA gene sequences with Pseudomonas lutea from soil (Peix et al. 2004).

A total of 12 bacterial strains were isolated from U. decussata sample number 10, 8 of which from the exterior part and 4 from the inner; among them, 6 strains isolated from the external part and 3 from the inner showed 98% of similarity both with an unidentified endophytic bacterium (DQ339615) from plants living under the snow in the Alps (unpublished data) and an unidentified and uncultured bacterium from an indoor habitat (AM696980, unpublished data). Similarly to the isolates 12Uae1 and 12Uae2 from U. antarctica, the strain 10Ude8 from U. decussata matched with the same bacterium (AM696982, 98% similarity). From the same thallus, strain 10Udi4 was 99% similar to the actinomycete Knoellia sp. (DQ812538). The presence of the strain 10Ude3 in the external fraction of the thallus is remarkable. This strain shows 97% similarity with the psychrophilic, ionizing radiation-sensitive bacterium Deinococcus alpinitundrae recently isolated from alpine environments (Callegan et al. 2008). From sample number 11 of U. decussata, only one strain was isolated: 11Udi1 shows 98 and 97% similarity with an unknown species (Bacterium H25 AY345548) and Paenibacillus sp. AJ011321, respectively. Three strains from the external part of the thallus of X. elegans and four from the interior of the lichen thallus were studied; all of them showed 99% similarity with G+ and G− of the genera Pseudomonas, Bacillus, Mycobacterium and Paenibacillus. The strain 14Xei4 showed 99% similarity with Mycobacterium sp. strain K128W (DQ372729), whereas the isolate 14Xei3 was 99% similar to both a Pseudomonas species (EF157292) and the unidentified bacterial strain M3C4 (AF479376) from glacial environments (Christner 2002).

The phylogenetic analysis generated a well-resolved tree in which most of the groups are supported by bootstrap values up to 97–100% (Fig. 1). The Phylum Verrucomicrobia was arbitrarily chosen as outgroup. All the strains studied belonged to four different Phyla: Actinobacteria as the most frequently represented one, as well as Firmicutes, Proteobacteria and Deinococcus-Thermus. Among the strains studied, many different phylotypes can be distinguished. One of the main clades (Clade A) in the Actinobacteria included Cluster 1 and Cluster 2, with bacteria isolated from lichens only. Isolates from the same lichen thallus with identical genotypes were not included in the tree. The first one included six strains isolated from U. decussata only. These strains differ from each other just in a few nucleotide positions, except for strain 10Ude7, which was in a more isolated position. Strains 10Ude1 and 10Udi1, genotypically identical to 10Ude2, were not included in the tree. These strains were phylogenetically related with an endophytic bacterial strain associated with subnival plants in alpine environments (DQ339615). Cluster 2 included strain 12Uae1 (from U. antarctica), which is identical to 12Uae2 from the same lichen, and to 10Ude8 from U. decussata. One of the closest neighbours was the new genus and species Subtercola frigoramans isolated from boreal groundwater (Mannisto et al. 2000). Based on their phylogenetic relationships, all these strains can be determined with certainty only at the family level as Microbacteriaceae. The second clade (Clade B), family Intrasporangiaceae, included only the strain 10Udi4 from U. decussata; the strain was in a group that comprised only Knoellia species and can therefore be assigned to that genus. Clade C is a well-supported clade (99.7% bootstrap support) including Arthrobacter species only. The only lichen strain in the group was the one isolated from U. antarctica (12Uae3), which thus can be assigned with certainty to the genus Arthrobacter. This clade includes strains with interesting biology and ecology, being psychrophilic or isolated from Antarctica (Loveland-Curtze et al. 1999). Clade D, family Nakamurellaceae, including the well-separated Cluster 3, is formed by strains from lichens only. The closest neighbour appears to be the new genus and species from rock, S. lacteus (Lee et al. 2008). The isolates 5Lfe2 and 5Lfe3 from Lecanora fuscobrunnea were not included in the tree because they are identical to 5LFe4 from the same lichen. The strain 14Xei4 (Clade E) was the only one in the Mycobacteriaceae: the closest relationship seems to be with the genus Mycobacterium.
https://static-content.springer.com/image/art%3A10.1007%2Fs00300-009-0686-2/MediaObjects/300_2009_686_Fig1_HTML.gif
Fig. 1

Molecular phylogeny based on 16S rRNA gene sequences indicating the positions of the bacterial isolates. This is an ML tree, based on 84 sequences and 1,519 positions, which has been generated using GTR + I + G model and calculated using ML in MrAIC software. Bootstrap values from 1,000 resampled data sets are shown

Four strains (11Udi1 from U. decussata, and 14Xei1, 14Xei2 and 14Xee1 from X. elegans) belong to the phylum Firmicutes. Our tree suggests that the isolates 11Udi1 and 14Xei2 belong to the genus Paenibacillus, the first one with closer phylogenetic relationships to Paenibacillus strains from Finland (Elo et al. 2001), and the second with a very close relationship to a psychrotolerant Paenibacillus species from Alaska. The subclade with the strain 11Udi1 includes the new species P. borealis AJ011327 (HM31), an atypical member of the same species AJ011325 (KN25), and a possibly new Paenibacillus species AJ011325 (KM8) (Elo et al. 2001). The strain 14Xei1, in a group purely constituted by entities belonging to the species Bacillus simplex, can be assigned to the same species. However, isolate 14Xee1, which is found in a well-separated group together with the halotolerant Bacillus sp. AY553105 from the Great Salt Plains of Oklahoma (Caton et al. 2004) and B. licheniformis AJ582722, can be assigned, at generic level only, to the genus Bacillus.

Strains 14Xee3, 12Uae4, 12Uai1 and 14Xei3 all belong to the genus Pseudomonas (Proteobacteria; Gammaproteobacteria; Pseudomonadaceae). 14Xee2 is genotypically identical to 14Xee3 from the same lichen and was therefore not included in the tree. The two isolates 12Uae4 and 12Uai1 from U. antarctica grouped together (Cluster 4) in a separated position with respect to their closest neighbours. They probably belong to a not yet described species. Strain 14Xee3 from X. elegans grouped together with different species of Pseudomonas and the last strain included here, 14Xei3, forms a distinct clade together with an isolate from sub-glacial environment.

Finally, the strain 10Ude3, the only one in the Phylum Deinococcus–Thermus, probably belongs to a not yet described species.

Discussion

This contribution represents the first analysis of culturable bacteria associated with lichens from the ice-free areas of Northern and Southern Victoria land of the Continental Antarctic region, including the McMurdo Dry Valleys. The coastal sites, where the climatic conditions are more favourable, host a higher lichen diversity, while few species colonize the slopes and inland areas (Øvstedal and Lewis Smith 2001). The McMurdo Dry Valleys are among the most hostile terrestrial environments of our planet, supporting only small, extremotolerant lichens. Most species are restricted to sheltered microniches, such as rock crevices. L. fuscobrunnea is among the rare species able to colonize exposed rock surfaces and to develop typical lichen morphology under these extreme conditions (Onofri et al. 2007). Other lichens lose their typical anatomical stratification in functional layers and rather colonize sub-surface fissures in the rocks. In our study, we focused on lichens that were phenotypically recognized by their typical morphology.

The same bacterial genotypes were often isolated from the surface and interior of the lichen thalli, suggesting that bacteria do not clearly distinguish between these locations. We generally have studied only distinct bacterial phenotypes from each lichen thallus. Therefore, it is possible that we miss a larger number of genetically distinct strains, which are morphologically similar. The isolation procedures failed for many of the lichen samples tested, but a total absence of culturable bacteria on these thalli is unlikely. The higher numbers of isolates have been obtained from lichens U. decussata from Vegetation Island and X. elegans from Kay Island, where the conditions are milder than in inland sites of continental Antarctica. This may indicate that the colonization of lichen thalli under harsher conditions is more selective. Strains may be more specific to lichens and culture conditions likely did not match these conditions in all cases. Furthermore, we are convinced that culture-independent approaches will discover a greater bacterial diversity, and perhaps also those strains with high specificity for lichen symbioses. Nevertheless, our molecular analyses revealed a considerable number of bacteria in these extreme habitats. Several genera, representing four bacterial phyla, were found (Fig. 1). Some of them have previously been found in association with fungi, e.g. Paenibacillus, Bacillus (Cardinale et al. 2006) and Pseudomonas (Barbieri et al. 2005; Warmink et al. 2008). Burkholderia and Azotobacter have not been detected in this study, although they were reported in older Russian literature as associated with lichens (see references in Cardinale et al. 2006). The finding of the genus Deinococcus, which has never been reported as associated with epilithic lichens is remarkable. Several 16S rRNA gene sequences and cultured examples of Deinococcus-like organisms were previously obtained from Antarctic samples of snow, McMurdo Dry Valley’s soil and cryptoendolithic communities (Siebert and Hirsch 1988; Carpenter et al. 2000; de la Torre et al. 2003; Hirsch et al. 2004; Shravage et al. 2007). Members of the Deinococcaceae are best known for their capacity to withstand and limit the DNA damage due to exposition to extremely large amounts of ionizing radiations (Battista et al. 1999). Nonetheless, some authors suggested this high resistance to be a consequence of an adaptation to repair DNA damage induced by desiccation (Mattimore and Battista 1996). Such tolerance to radiation is an advantage for organisms colonizing one of the driest deserts at high latitude. From a phylogenetic point of view, our Deinococcus isolate 10Ude3 is only distantly related with the others already reported from the Antarctic, and likely represents a new species. The thermophilic Deinococcus murrayi from hot springs (Ferreira et al. 1997) is in a basal phylogenetic position to this group. We included in our comparison the sequences of two Deinococcus species from the South Pole AF239213 and AF239800 (Carpenter et al. 2000); however, the sequences had too little overlap with ours to get reliable alignments for phylogenetic reconstruction. Anyway, in a separate comparison, the sequences of these strains appeared to be very distinct from all the others, including ours (data not shown). The strain 10Ude3 was also not closely related with the uncultured Deinococcus AY250871 from Antarctic cryptoendolithic communities (de la Torre et al. 2003), which was in a basal position of the Deinococcus/Thermus clade (Fig. 1). The closest neighbour was instead the new psychrophilic species Deinococcus alpinitundrae with similar ecology. Strain 10Ude3 also seems to be dependent on cold environments and was unable to grow above 25°C (Table 3).

Judging from data reported in Table 2 and in the tree, individual lichen thalli may harbour a considerable bacterial biodiversity. Nonetheless, closely related genotypes are recurrent in the same thallus and often grouped together in the tree forming separated and well-supported clusters (Cluster 1, 3, 4; Fig. 1). Only one of such clusters included strains isolated from L. fuscobrunnea (Cluster 3). These bacterial genotypes grouped with some unidentified Actinomycetales bacteria (98% similarity) and with the new genus and species S. lacteus, known from rocks (Lee et al. 2008).

The phylogenetic analysis highlighted relationships of many of the isolates with bacteria present worldwide even in temperate climates; thus these do not seem to be specialized in colonizing lichen thalli or Antarctic environments. Similarly, different studies performed on lichens from milder zones revealed the presence of bacterial species typically occurring on other substrata than lichens. Bacteria isolated fall in the same taxonomic groups already reported in this previous investigation (Cardinale et al. 2006) with respect to high taxonomic ranks, but with remarkable differences at the generic level. For example, we did not find strains of Burkholderia, which seem common in temperate lichens. On the other hand, Deinococcus is present in Antarctic lichens, and in the Actinobacteria, we found species related to the genera Knoellia, Mycobacterium and Arthrobacter, while lichens of temperate climates rather contained Streptomyces or Micromonospora.

Still, remarkable are the phylogenetic relations, with many bacteria living under different extreme conditions. Clusters 1 and 2, for instance, are related to bacteria living in very cold environment: an endophytic bacterium (DQ339615) associated with alpine subnival plants and one (AF224723) isolated from boreal groundwater, respectively. The strain 11Udi1 grouped together with many strains of Paenibacillus borealis from Finland (Elo et al. 2001), whereas strain 14Xei2 had as closest neighbour a psychrotolerant Paenibacillus species from Alaskan tundra. The strain 14Xee1 is related to the halotolerant Bacillus sp. (AY553105) isolated from the Great Salt Plains of Oklahoma (Caton et al. 2004) and, finally, strain 14Xei3 grouped together with the glacial ice bacterium AF479376 isolated from glacial environments (Christner 2002). The adaptation to the cold and the ability to tolerate water deficiency are closely related abilities, and water loss is one of the main constraints that both cold- and halotolerant organisms have to face (Ruisi et al. 2007). Therefore, in this study phylogeny seems be predictive of psychrophily and desiccation tolerance. Even in the genus Pseudomonas, included in the tree, the species P. putida is known for its anhydrobiotic capacity (García de Castro et al. 2000). Most lichen species are desiccation-tolerant organisms (Kranner et al. 2008) and their success, particularly, in Antarctic environments is strictly connected with this ability. We were therefore not surprised that microbes associated with thalli of Antarctic lichens have similar ecological capacities.

Low temperatures and severe water deficiency are two of the main limiting factors for life in the ice-free areas of the Antarctic continent. There and in sub-antarctic islands, we expect that the bacterial strains cultured from lichens are present also in other niches, such as mosses, plant debris or in microbial consortia. Rapid thermal fluctuations rather than low temperatures can be the main limitations affecting microbial colonization of lithic substrates: temperature fluctuates over a range of about 8°C, crossing the freezing point up to 14 times in a 42-min period and rock surfaces can reach as much as 20°C above air temperature (McKay and Friedmann 1985). Still, as response to extreme conditions, Antarctic lichens are generally dark or even black; the colour improves their thermal capacity and, as a result of shelter and sun exposure, on moist lichen thalli temperatures up to 10°C have been recorded (Kappen 1993). Considering the range and the optimum temperatures for psychrophilic and psychrotolerant bacteria (Morita 1975), we selected 10 and 25°C as isolating temperatures. None of our strains could be classified as truly psychrophilic according to the definition by Morita (1975), because all of them were still able to grow at 25°C, although most of them were unable to grow at higher temperatures (Table 3). The predominance of psychrotolerant, rather than psychrophilic bacteria in the Antarctic, was previously reported by Helmke and Weyland (2004).

The ability of some of the isolated strains to grow very well in a rather wide range of temperatures is remarkable: this has been already suggested for other Antarctic microorganisms as an adaptation to an environment characterized by very wide thermal fluctuations (Zucconi et al. 1996; Selbmann et al. 2005). Judging from data reported in Table 3, very often not strictly related genotypes isolated from the same thallus showed different temperatures relations. For instance, strains in Cluster 1 and from U. decussata, have more restricted range compared with strains 10Udi4 and 10Ude3 from the same thallus; strains 14Xee2, 14Xei2, 14Xee3 and 14Xei1 from X. elegans appear to be rather psychrotolerant and able to grow in a wider range of temperatures compared with strain 14Xee1, whose preferences seem to be at rather higher temperatures. We assume that these different adaptations may allow the coexistence of considerable bacterial biodiversity in a single lichen thallus, which provides specific niches for different isolates.

The phylogenetic analysis revealed that many of the strains studied pooled in well-supported groups that only comprised bacteria from lichens. For example Clusters 1, 2, 3 and 4 may represent new species, but additional physiological studies are required before description of new taxa. Further work is also needed to investigate potential interactions of lichens and lichen-associated bacteria (e.g. Grube et al. 2009). It will be particularly interesting to analyse these strains for potential biotechnological exploitation and to reveal functional properties of culturable, cold-adapted bacteria. Bacterial communities could be involved in nutrient mobilization in lichens, in particular due to their lytic functions, their ability to synthesize essential amino acids, to fix atmospheric nitrogen or to solubilize minerals, including phosphates, from the rock substratum (Liba et al. 2006; Cardinale et al. 2008; Grube et al. 2009). An artificial co-culture study of lichen mycobionts with nitrogen-fixing bacteria indicated increased capacity of rock weathering with bacterial symbionts (Seneviratne and Indrasena 2006), suggesting that colonization and establishment of lichens on rocks could be facilitated by bacterial actions. Lytic activities were detected in culturable, lichen-associated bacteria from temperate habitats (Grube et al. 2009). In addition to enzymatic capacities, bacteria from these habitats may be producers of new compounds. Actinobacteria isolated from lichens have already been successfully screened for such compounds (Davies et al. 2005; Williams et al. 2008).

Microscopic studies indicated that lichens are a rather selective environment, in which only specific bacteria can grow (Cardinale et al. 2008). This might also be attributed to the diverse secondary metabolites produced by lichen-forming fungi in symbiotic stages. Some of these compounds, e.g. usnic acid, can influence bacterial growth (e.g. Boustie and Grube 2005; Francolini et al. 2004). A recent study that includes culture-independent approaches showed that lichens in the same habitat harbour species-specific bacterial communities (Grube et al. 2009). Data on phylogenetic relationships and physiological profiles reported here indicate that the environmental constraints are one of the prime factors of at least the culturable bacterial fraction from Antarctic lichens. Thus, species of lichens occurring in different climatic regions might well harbour bacterial communities that are markedly distinct from those reported here. In this respect, it will be interesting to compare lichen species, which have wide ecological amplitudes and which occur also in other regions besides Antarctica.

Acknowledgments

The authors would like to thank PNRA (Italian National Program for Antarctic Research) for supporting sample collection and Italian National Antarctic Museum “Felice Ippolito” for supporting CCFEE (Culture Collection of Fungi From Extreme Environments).

Copyright information

© Springer-Verlag 2009